Enhancing Learning Performance through LLM-Driven Adaptive Contents Based on Facial Engagement Detection

Recent advancements in AI have opened new avenues for personalized education, enabling learning environments to adapt to individual needs. This study explores the intersection of AI and Image Processing, mainly using facial emotion detection to monitor cognitive engagement and distraction during learning tasks. Previous research highlights the efficacy of emotion detection in identifying cognitive states and improving educational outcomes through personalized learning environments. Our primary goal is to evaluate how AI can adapt educational content in real time to enhance learning effectiveness based on emotional feedback. We utilized an LLM to develop educational content and identified moments of distraction or disengagement by monitoring learners' facial emotions while interacting with the system. Upon detecting distraction, the LLM dynamically restructures the content, subtly reintegrating it into subsequent materials without the learners' awareness. This method reinforces learning by presenting material in varied contexts, thereby improving learners' understanding. Our preliminary results indicate that this approach enhances comprehension levels in distracted learners. Furthermore, integrating emotion-based monitoring with LLMs may provide a responsive educational framework that adapts to learners' immediate needs, promoting deep cognitive engagement and a pleasant personalized learning experience. These findings suggest broader applications beyond academic learning, including professional training. It is the ambition of any learning system to be able to detect any attention fall and to have smart reactions without learners' awareness, which is the vision of our approach.